I look at SMI-S
as a standard
that is actually going to be implemented by many of the companies in the storage industry in 2004. This is a very different discussion then the standards that have come and gone (Fibre Alliance MIB, WideSky and many others). What SMI-S has that the others didn't is a critical mass of support and vendors actually partnering to deliver a specification. The specification is relatively easy to implement and can be done either alone or with the help of third parties. AppIQ is offering an integration kit for those companies that want to go it on their own but don't have the time or resources to integrate from scratch or AppIQ will do the full integration for a fee. (We call that Non-Recurring Engineering or NRE).
I am very excited about the prospects that SMI-S can bring but also understand that one must proceed with caution because solutions are not made by just one company. Solutions are made by combining technologies from multiple companies all working together and within enterprise IT shops. If one of the pieces of the solution puzzle doesn't support SMI-S than one is relegated to using a shim like SNMP or a vendor's proprietary APIs to create a full solution.
This isn't very different than the early days of SAN management
when SNMP was the basis for management and all of the switch vendors compiled a standard MIB with proprietary extensions but then an IT manager wanted to understand what the host HBAs were doing so he found a proprietary API to integrate with from the HBA vendors since the HBAs didn't have support in the Host Resource MIB ... OK, I am going down a rat hole. What this all means is that a standard is only as good as the vendors supporting it and with SMI-S the big guys and the small guys alike are behind it 100%.
Let me know if you have any questions or thoughts.
Brett P. Cooper
Editor's note: Do you agree with this expert's response? If you have more to share, post it in one of our
This was first published in February 2004